def get_activated_node(self):
"""returns the node with the highest activation
"""
activations = []
for i in self.nodes:
activations.append(i.activation)
return self.nodes[auks.posMax(activations)]
def get_closest_concept_tagsim(self, object_data):
""" returns the tag and similarity of the closest concept to the given stimulus
"""
similarities = []
for i in self.concepts:
similarities.append(auks.calculate_similarity(object_data, i.get_data(), cfg.sensitivity))
max_pos = auks.posMax(similarities)
return (self.concepts[max_pos].tag, similarities[max_pos])
def get_h_tag(self, v_tag):
""" retrieves the h_tag with the highest association for the given v_tag
if there are more than one h_tags with the highest association value,
the first one will be returned
"""
v_index = -1
for count, i in enumerate(self.vertical_tags):
if i == v_tag:
v_index = count
break
if v_index == -1:
return "label_unknown"
else:
max = auks.posMax(self.matrix[v_index])
return self.horizontal_tags[max]
def get_v_tag(self, h_tag, inaccuracy=None):
""" retrieves the v_tag with the highest association for the given h_tag
if there are more than one v_tags with the highest association value,
the first one will be returned
if inaccuracy == True/1, the v_tag with the 2nd highest association is returned
"""
h_index = -1
for count, i in enumerate(self.horizontal_tags):
if i == h_tag:
h_index = count
break
if h_index == -1:
return "tag_unknown"
else:
value_list = []
for i in self.matrix:
for count, j in enumerate(i):
if count == h_index:
value_list.append(j)
if inaccuracy:
return self.vertical_tags[auks.posSemiMax(value_list)]
else:
return self.vertical_tags[auks.posMax(value_list)]
